01.02.2015 Views

Article - The New IHO S-102 Standard - Caris

Article - The New IHO S-102 Standard - Caris

Article - The New IHO S-102 Standard - Caris

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

<strong>The</strong> <strong>New</strong> <strong>IHO</strong> S-<strong>102</strong><br />

<strong>Standard</strong><br />

Charting a <strong>New</strong> Frontier for Bathymetry<br />

Soundings and contours are the only official way data producers can push bathymetric information to the wide<br />

hydrographic community. <strong>The</strong> introduction of the S-<strong>102</strong> standard for bathymetry will enable many possibilities within the<br />

community of bathymetry users. Liaising with the International Hydrographic Organization’s (<strong>IHO</strong>) TSMAD working group,<br />

the Canadian Hydrographic Service (CHS) and the US Naval Oceanographic Office (NAVO) led the development and<br />

practical testing of this revolutionary standard, officially called the Bathymetric Surface Product Specification S <strong>102</strong>. In<br />

partnership with IIC Technologies, GeoNet Technologies, and CARIS, CHS created one of the first prototypes of S <strong>102</strong><br />

datasets, using the CARIS Bathy DataBASE software suite. A portfolio of 86 high-definition bathymetric charts was<br />

subsequently produced within a few weeks, successfully validating a specification that will potentially change the<br />

manipulation of bathymetric data we’ve known for years.<br />

Marc Journault<br />

Canadian<br />

Hydrographic<br />

Service<br />

Louis Maltais<br />

Canadian<br />

Hydrographic<br />

Service<br />

Ed Kuwalek<br />

IIC Technologies,<br />

Canada<br />

In shallow-water, ships navigation<br />

and manoeuvring commands precautions<br />

and decisions made with<br />

the best information available and<br />

usable in a secure manner. In these<br />

circumstances, the most recent<br />

details of the bottom shape are usually<br />

considered essential. A three<br />

dimensional ‘picture’ of the bottom<br />

would meet this requirement,<br />

and this is exactly what is meant by<br />

High-Definition Gridded Bathymetry<br />

(HDGB).<br />

<strong>The</strong> grid bathymetry can be<br />

described as a Navigation Surface<br />

(developed by Shep Smith) or a Digital<br />

Terrain Model (DTM) of the<br />

seafloor in the form of regular rectangular<br />

meshes. By its nature, the<br />

resolution or the density of bathymetric<br />

data collected usually varies<br />

according to the depth range, and it<br />

would be arguable to fix a quantitative<br />

value to the term high definition<br />

for bathymetry. In spite of this, the<br />

concept of HDGB used in this article<br />

is simply defined as a much higher<br />

resolution than what is available on<br />

the charts. For the source bathymetry<br />

coming from a multi-beam system<br />

the resolution can be similar to<br />

the insonified footprint. For digitised<br />

legacy data this can be the equivalent<br />

of a field sheet density.<br />

<strong>The</strong> existing tools to process and<br />

manage bathymetric data generate<br />

and combine grid datasets that are<br />

used for chart compilation and contour<br />

generation. Such grid datasets<br />

of bathymetry can also be a product<br />

by itself, created for different usage<br />

like hydrodynamic modelling, spatial<br />

data analysis with GIS, coastal management<br />

with land DTM, or marine<br />

navigation.<br />

To develop a HDGB product for<br />

navigation CHS adopted a collaborative<br />

approach where navigators<br />

and software manufacturers were<br />

informed of the intentions and consulted<br />

for specifications of a future<br />

product. For the consultation, prototype<br />

datasets were produced and<br />

made available in a simple format<br />

for experimentation and trials. <strong>The</strong><br />

work of the <strong>IHO</strong> Transfer <strong>Standard</strong><br />

Maintenance and Application Development<br />

working group (TSMAD)<br />

on the new Geospatial <strong>Standard</strong><br />

for Hydrographic Data (S-100) was<br />

promising and CHS decided to use<br />

it. S-100 is a framework standard for<br />

the registration, maintenance and<br />

capture of hydrographic geospatial<br />

data and product specifications. It is<br />

based on the international ISO 19000<br />

series of geographic standards, and<br />

it is flexible. CHS and NAVO worked<br />

together to draft and propose a product<br />

specification for grid bathymetry<br />

based on previous work done by the<br />

Open Navigation Surface group. <strong>The</strong><br />

standard is named Bathymetric Surface<br />

Product Specification and identified<br />

as S-<strong>102</strong>.<br />

At the 3 rd HSSC meeting in November<br />

2011, the <strong>IHO</strong> TSMAD invited<br />

the Hydrographic Services and<br />

<strong>Standard</strong>s Committee (HSSC) to<br />

approve the final draft of S-<strong>102</strong> and<br />

instruct the IHB to submit it to<br />

Member States for their endorsement.<br />

Subsequently, the <strong>IHO</strong> circular<br />

Hydro international | may 2012 | 21


Figure 1: Example<br />

of a level 2<br />

(coastal) tiling<br />

scheme for the<br />

St.Lawrence River.<br />

letter 10/2012 requested the Member<br />

States to review and consider the<br />

draft edition of S-<strong>102</strong> which is available<br />

on the <strong>IHO</strong> website.<br />

To partition HDGB coverage, to allow<br />

unambiguous dataset exchange and<br />

to facilitate the updates, CHS intends<br />

to implement a systematic tiling<br />

scheme with three levels of resolution.<br />

Each tile is comprised of 1000 by<br />

1000 grid cells, the level 1 (harbour)<br />

is 0.02º x 0.02º, the level 2 (coastal) is<br />

0.1º x 0.1º, and the level 3 (overview) is<br />

1º x 1º. For the low and mid latitudes,<br />

the orientation of the tiles fit the<br />

version of the <strong>IHO</strong> S-<strong>102</strong> product<br />

specification. Developing a uniform,<br />

high-definition bathymetric surface<br />

for the St. Lawrence River between<br />

ports of Quebec and Montreal was the<br />

key goal for the project. In addition, as<br />

this was the first attempt to produce<br />

a large number of datasets based on<br />

this new product specification, the<br />

project provided an ideal opportunity<br />

to validate the feasibility of S-<strong>102</strong> production<br />

in practice.<br />

<strong>The</strong> main requirements for the portfolio<br />

included creating seamless<br />

data coverage with a final surface<br />

S-100 framework standard<br />

represents a step foreword for<br />

electronic navigation products<br />

and services<br />

meridians and parallels with an origin<br />

based on a round number in latitude<br />

and longitude. Figure 1 shows an<br />

example of the tiling scheme.<br />

Production of Datasets for Trials<br />

In January 2011, CHS contracted<br />

IIC Technologies and GeoNet Technologies<br />

to produce a portfolio of<br />

86 HDGB datasets based on a draft<br />

grid resolution of 0.0001º (approximately<br />

8 metres) divided into 0.1º x<br />

0.1º tiled datasets. <strong>The</strong> final bathymetric<br />

surface needed to be created<br />

by integrating a variety of source<br />

datasets available for the area, ranging<br />

from sparse sounding sets to<br />

much higher resolution grid data<br />

for channels, while also using ENCderived<br />

high-water lines along with<br />

the corresponding drying spot values.<br />

<strong>The</strong> intention behind using all<br />

available source data was to integrate<br />

all information in a complete dataset,<br />

capable of significantly enhancing<br />

the information already available<br />

on the ENC cells. Similarly the project<br />

required the deliverables to be provided<br />

in multiple formats including<br />

CSAR (CARIS Spatial Archive), BAG<br />

and 32-bit GeoTiff files to cater to a<br />

variety of potential end-users.<br />

<strong>The</strong> production processes consisted<br />

of four main stages: source data<br />

assessment and preparation, point<br />

data processing, surface data integration<br />

and conflict resolution, and<br />

final product surface extraction<br />

and export. During the first stage all<br />

source data was assessed and prepared<br />

for production. CARIS S-57<br />

Composer and BASE Editor were<br />

used to create all auxiliary components,<br />

such as the tiling scheme, coverage,<br />

and breakline vector layers. <strong>The</strong><br />

same tools were also used to cut all<br />

data sources into individual data tiles<br />

to facilitate their efficient processing.<br />

During stage two the XYZ point<br />

data was loaded into BASE Editor<br />

and used to build a TIN model which<br />

was enhanced by the addition of the<br />

high-water-based breakline vector<br />

layer. Subsequently, a high-resolution<br />

bathymetric surface was interpolated<br />

from the TIN models for each dataset.<br />

During stage three the additional<br />

22 | may 2012 | Hydro international


which will be identified as mandatory,<br />

conditional or optional, will be<br />

selected from these ISO 19115 metadata<br />

packages.<br />

Figure 2: S-<strong>102</strong> Production Process.<br />

high-definition surface data available<br />

for channels was merged with<br />

the product surface data created in<br />

stage two. This was achieved by using<br />

rule-based conflict resolution tool<br />

available in BASE Editor. In the final<br />

stage the coverage polygon was used<br />

to extract the final product surface,<br />

effectively facilitating a complete<br />

river bank to river bank data coverage.<br />

<strong>The</strong>reafter BASE Editor was used<br />

to export CSAR and BAG deliverables<br />

and CARIS Easy View was used<br />

to export 32-bit GeoTiff rasters. Figure<br />

2 illustrates the key stages of the<br />

data production process used during<br />

the project.<br />

All project goals were successfully<br />

achieved and the required portfolio<br />

of 86 HDGB datasets was efficiently<br />

produced, confirming the feasibility<br />

of HDGB data production using the<br />

currently available CARIS production<br />

tools.<br />

Generation of S-<strong>102</strong><br />

Bathymetry Grids<br />

As described above, the interpolated<br />

grids from the point data are combined<br />

with the high-definition grids<br />

from the ship channel to produce the<br />

final grids for each cell, in the CSAR<br />

format, which are then exported to<br />

the desired carrier format of the grid.<br />

<strong>The</strong> draft S-<strong>102</strong> standard follows the<br />

principle of separating the ‘carrier’<br />

from the ‘content’ which means that<br />

the encoding of the grids will be flexible.<br />

Following that principle, the<br />

CARIS software tools provide choices<br />

for export formats. At the time of this<br />

S-<strong>102</strong> prototype work, the choices<br />

were:<br />

• 32bit Floating Point GeoTIFF format.<br />

This option for a carrier format<br />

is available from the GDAL<br />

open source library ( 1). <strong>The</strong><br />

32bit floating point option is used<br />

so that the depth values can be<br />

encoded without loss of precision.<br />

• BAG format. At the time, the<br />

CARIS Bathy DataBASE (BDB) v3.1<br />

supported BAG v1.3. <strong>The</strong> support<br />

for BAG format is possible through<br />

the API provided by the Open Navigation<br />

Surface Working Group (<br />

2) of which CARIS is a participant.<br />

Since then, BDB has been updated<br />

to support BAG v1.4. CARIS is currently<br />

working with the ONSWG on<br />

BAG v1.5, which will be supported<br />

when finalised and released.<br />

Metadata<br />

Adjusting this production line to<br />

match the coming S-<strong>102</strong> standard is<br />

going to require the creation of metadata<br />

that is compliant with the S-<strong>102</strong><br />

profile. <strong>The</strong> details of the S <strong>102</strong> metadata<br />

profile are still being developed<br />

but many aspects of the content are<br />

already known. Fortunately, this area<br />

has received significant attention<br />

over the past two years at CARIS with<br />

the creation of the metadata profile<br />

for the CSAR format that is compliant<br />

with the ISO 19115 / 19115-2 standards<br />

and using XML encoding from<br />

ISO 19139. Drafts of the S-<strong>102</strong> standard<br />

indicate that its metadata profile<br />

will share many of the same elements<br />

that are already being populated in<br />

exported CSAR metadata XML files.<br />

<strong>The</strong> drafts of S-<strong>102</strong> metadata profile<br />

indicate that most of the core metadata<br />

elements in the profile, each of<br />

Evolution of CARIS BDB and<br />

S-<strong>102</strong> Software Tools<br />

In the period since these prototype<br />

S-<strong>102</strong> gridded products were produced,<br />

the software tools have evolved<br />

which could improve the workflow<br />

even further. For example, the Warp<br />

function, which has been available<br />

since BDB 3.2, can transform a gridded<br />

surface to a new coordinate system<br />

and/or re-sample the surface<br />

onto a new grid resolution. This could<br />

save steps in turning what started out<br />

as projected bathymetry in grid resolutions<br />

of metres on the ground, into<br />

unprojected grids at the appropriate<br />

decimal degree resolutions with one<br />

function.<br />

<strong>The</strong> next step in the evolution of software<br />

tools for S-<strong>102</strong> production will<br />

be the creation of a new Export to<br />

S-<strong>102</strong> procedure that addresses the<br />

following all in one function:<br />

• Take a CSAR grid as input which<br />

has been prepared with the correct<br />

coordinate system and grid<br />

resolution.<br />

• Select an option for the output ‘carrier’<br />

format of the gridded bathymetry.<br />

(e.g. GeoTIFF, BAG, and others<br />

as required)<br />

• Optionally apply a tiling scheme<br />

which will cut the input grid into<br />

the desired tiles of specified width/<br />

height in decimal degrees.<br />

• Apply a defined naming convention<br />

to the output tiled files.<br />

• Generate an accompanying XML<br />

metadata. This metadata could<br />

then be used by data producers to<br />

populate a metadata catalogue web<br />

service<br />

S-<strong>102</strong> Gridded Bathymetry<br />

on the Web<br />

<strong>The</strong> CARIS web mapping solution,<br />

Spatial Fusion Enterprise (SFE), supports<br />

the discovery and dissemination<br />

of high-definition bathymetry<br />

through the Internet using OGC,<br />

Open Geospatial Consortium ( 3),<br />

standard formats. S-<strong>102</strong> bathymetric<br />

datasets residing in the CARIS BDB<br />

database as CSAR grids are requested<br />

by SFE through a specialised bathymetry<br />

web service. <strong>The</strong> CSAR format<br />

was designed with the optimisation<br />

Hydro international | may 2012 | 23


Figure 3: Filtered ENC and S-<strong>102</strong> in background.<br />

required for web distribution of this<br />

type of data in mind. OGC requests<br />

are made from any OGC capable client.<br />

<strong>The</strong> SFE Server handles the OGC<br />

request and returns the relevant information<br />

residing in the bathymetry<br />

database. SFE currently supports the<br />

OGC WMS, WMTS or WCS services<br />

for bathymetry data: it is the WCS<br />

(Web Coverage Service) that provides<br />

download capabilities for the grids<br />

including options for the format to be<br />

received from the download. Efficient<br />

browsing and simple query capabilities<br />

are increasingly being provided<br />

through the use of WMTS (Web Map<br />

Tile Service) which is a variant on<br />

the traditional WMS (Web Map Service)<br />

protocol which chunks data into<br />

image tiles. <strong>The</strong>se tiles are cached on<br />

initial draw meaning that the next<br />

time an image request is made the tile<br />

is drawn from cache rather than by<br />

going back to the database.<br />

Use of S-<strong>102</strong> for navigation<br />

With S-<strong>102</strong> it will be easy for hydrographers<br />

to provide navigators with<br />

HDGB datasets very rapidly to enrich<br />

and update the information of their<br />

ENCs. Partitioned into standardised<br />

regular geographic cells, the provision<br />

of updates becomes easy by just supplying<br />

an entire new cell dataset with<br />

the new bathymetry to the users who<br />

then just supersedes the tile.<br />

A possible way to use HDGB for navigation<br />

is to put the S-<strong>102</strong> data in the<br />

background with an overlap of S-101<br />

without the display of S-101 soundings,<br />

depth areas and contours.<br />

Figure 3 shows such usage.<br />

But electronic chart systems manufacturers<br />

can do more than just displaying<br />

it. <strong>The</strong> HDGB can be used: to<br />

generate safety contour on the fly; to<br />

compute the available water column<br />

by the addition of water level (tidal<br />

data) on top of the bathymetry; to calculate<br />

a dynamic under keel clearance;<br />

etc.<br />

<strong>The</strong> exploitation of HDGB in an innovative<br />

way is left to the users and the<br />

software industry. <strong>The</strong> CHS and its<br />

partners wish to involve them at an<br />

early stage of design to consider feedback<br />

from the users’ communities.<br />

Conclusion<br />

<strong>The</strong> <strong>IHO</strong> S-100 framework standard<br />

represents a step foreword for<br />

electronic navigation products and<br />

services. This flexible standard is a<br />

building block to rethink the way<br />

hydrographic organisations (HO) can<br />

deliver their official data and services.<br />

This S-<strong>102</strong> experiment clearly showed<br />

how modern tools and standards will<br />

enable quick delivery of rich bathymetric<br />

information. It’s time to take<br />

a step ahead and test these datasets<br />

with users. This year the focus will be<br />

on getting the ECS and ECDIS manufacturers<br />

involved in exploring the<br />

possibilities of S-<strong>102</strong>.<br />

<strong>The</strong> Authors<br />

Marc Journault has been working for<br />

the Canadian Hydrographic Service for 28<br />

years. He is manager of the Quebec region’s<br />

Marine Geomatics Division, responsible for<br />

technological support and development in<br />

the fields of hydrographic surveys, nautical<br />

publications and marine services.<br />

Louis Maltais has been working for the<br />

Canadian Hydrographic Service (CHS) for 13<br />

years. He is involved in numerous R&D projects,<br />

implementing solutions to improve productivity,<br />

efficiencies and to better meet the clients’<br />

needs.<br />

Ed Kuwalek is the director of R&D at IIC<br />

Technologies in North Vancouver, Canada. Mr.<br />

Kuwalek has 18 years of experience in the field<br />

of nautical charting, electronic-chart production<br />

systems and spatial data conversions. He has<br />

been involved in the technical leadership of<br />

many industry-leading projects and is an active<br />

participant in the <strong>IHO</strong> Transfer <strong>Standard</strong> and<br />

Maintenance Working Group.<br />

1. www.gdal.org<br />

2. www.opennavsurf.org<br />

3. www.opengeospatial.org<br />

Hydro international | may 2012 | 25

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!